List your company

Novel View Synthesis on Synthetic Scenes LDR-OE: t1, t3, t5 (test)

43.78PSNRGaussHDR (Scaffold-GS)
12.777620.826328.87536.9237Nov 29, 2021Jun 17, 2022Jan 3, 2023Jul 22, 2023Feb 7, 2024Aug 25, 2024Mar 13, 2025

Evaluation results

MethodLinks
GaussHDR (Scaffold-GS)Training setting=Exp-3, Representation=Scaffold-GS2025.0343.780.990.003
HDR-Scaffold-GSTraining setting=Exp-3, Representation=Scaffold-GS2025.0343.160.9890.004
Ours (Scaffold-GS)Scene Representation=Scaffold-GS, Training Setting=Exp-12025.0342.940.9880.004
GaussHDR (3DGS)Training setting=Exp-3, Representation=3DGS2025.0342.290.9850.007
HDR-Scaffold-GSScene Representation=Scaffold-GS, Training Setting=Exp-12025.0342.210.9860.005
Ours (3DGS)Scene Representation=3DGS, Training Setting=Exp-12025.0341.510.9840.008
HDR-GSTraining setting=Exp-3, Supervision (HDR GT)=false, Implementation=Re-implementation2025.0341.130.9830.009
HDR-GSTraining setting=Exp-3, Supervision (HDR GT)=true2025.0341.10.9820.011
HDR-NeRFTraining setting=Exp-3, Implementation=Re-implementation2025.0340.160.9760.019
HDR-GSScene Representation=3DGS, Training Setting=Exp-1, Implementation=Re-implemented2025.0339.350.9770.012
HDR-NeRF2021.1139.070.9730.026
HDR-NeRFScene Representation=NeRF, Training Setting=Exp-12025.0339.070.9730.026
NeRF-GTTraining data=LDR images with consistent exposures or HDR images, Role=Upper bound2021.1137.660.9650.028
NeRF-WNote=Exposures of input views are randomly selected from all five exposures to learn five appearance vectors for testing2021.1129.830.9360.047
NeRF2021.1113.970.5550.376